This paper describes the use of a video-based motion capture system to assess spinal mobility in patients with ankylosing spondylitis (AS). The aim of the study is to assess reliability of the system comparing it with conventional metrology in order to define and analyze new measurements that reflect better spinal mobility. A motion capture system (UCOTrack) was used to measure spinal mobility in forty AS patients and twenty healthy subjects with a marker set defining 33 3D measurements, some already being used in conventional metrology. Radiographic studies were scored using the modified Stoke Ankylosing Spondylitis Spine Score index (mSASSS). Test-retest reliability studies were performed on the same day and over a two-week period. Motion capture shows very high reliability with Intraclass Correlation Coefficient values ranging from 0.89 to 0.99, low Standard Error of the Measurement (0.37-1.33 cm and 1.58°-6.54°), correlating very well with the Bath Ankylosing Spondylitis Metrology Index (BASMI) (p < 0.001) and, in some individual measures (cervical flexion, cervical lateral flexion, back inclination, shoulder-hip angle and spinal rotation), with mSASSS (p < 0.01). mSASSS also added significantly to the variance in multivariate linear regression analysis to certain measures (back inclination, cervical flexion and cervical lateral flexion). Quantitative results obtained with motion capture system using the protocol defined show to be highly reliable in patients with AS. This technique could be a useful tool for assessing the outcome of the disease and for monitoring the evolution of spinal mobility in AS patients.